Actions

Ontolog Forum

Session Synthesis
Duration 1.5 hour
Date/Time Feb 28 2018 17:00 GMT
9:00am PST/12:00pm EST
5:00pm GMT/6:00pm CET
Convener KenBaclawski

Ontology Summit 2018 Synthesis Session 1

Agenda

The aim of this week's session is to identify the major issues and open problems of the sessions that have been held so far. Unlike previous Ontology Summit Synthesis sessions, the aim is not to summarize the subject matter or the lessons learned so far. Sources for the major issues and open problems include the invited speaker presentations, chat log comments, blog posts and video recordings.

Conference Call Information

  • Date: Wednesday, 28-February-2018
  • Start Time: 9:00am PST / 12:00pm EST / 6:00pm CET / 5:00pm GMT / 1700 UTC
  • Expected Call Duration: ~1.5 hours
  • Video Conference URL: https://bluejeans.com/703588230
    • If you have not used BlueJeans before, then connect to the URL above before the meeting time so that the required plug-in can be installed.
  • Chatroom: http://webconf.soaphub.org/conf/room/ontology_summit_2018
    • Instructions: once you got access to the page, click on the "settings" button, and identify yourself (by modifying the Name field from "anonymous" to your real name, like "JaneDoe").
    • You can indicate that you want to ask a question verbally by clicking on the "hand" button, and wait for the moderator to call on you; or, type and send your question into the chat window at the bottom of the screen.
  • This session, like all other Ontolog events, is open to the public. Information relating to this session is shared on this wiki page.
  • Please note that this session may be recorded, and if so, the audio archive is expected to be made available as open content, along with the proceedings of the call to our community membership and the public at-large under our prevailing open IPR policy.

Attendees

Proceedings

KenBaclawski: Welcome to the synthesis session.

David Whitten: Ken asks us to share current issues in this field.

David Whitten: Cory suggests that one issue/concern/gap is a lack of qualification of context by speakers. What Context consists of may be hinted but not explicit.

Gary Berg-Cross: Context - circumstances that form the setting for an Event,statement, or idea, and in terms of which it can be better understood and assessed.

David Whitten: Jack asks if we are trying to make an ontology of context ?

David Whitten: Cory disagrees and points out how context is defined in the terms of the details of properties.

David Whitten: Cory says as a general definition, at least it is a starting point.

David Whitten: I think Pat Hayes defined Context by a negative definition like 'garbage' is defined by what is not there.

David Whitten: Gary points out a distinction between the physical and the formal world. Reminiscent of the John Sowa's meaning triangle.

David Whitten: Jack points out that environment and context are conflated in the systems world. What outside a field or domain impinges on the central system.

David Whitten: Correction John Sowa; C.S. Pierce's triangle of meaning.

David Whitten: Cory says the context mediates between the things it actualizes.

David Whitten: Janet agrees with Jack's point that there is a difference between environment and context. Context might be focused only on text, with hermeneutics allowing for a systematic way to interpret a text.

Gary Berg-Cross: It would be meaningful to an Ontology Summit to discuss the issues that are the external implications for Ontologies looked at from various contexts as well as the internal one of how to routinely incorporate these ideas into the building of ontologies. This brings us to ontological engineering.

David Whitten: Todd feels there is a context of agents as well as of text.

David Whitten: Janet feels that an environment can influence whether the interpretants recognize it or not. Context usually is known to the interpretant.

Gary Berg-Cross: BTW, I have to go at 1 so I can drive after lunch to 1st U.S. Semantic Technologies Symposium in Fairborn, OH starting tomorrow. Anyone else going??

David Whitten: Jack says individual humans have different influence-ability by context.

David Whitten: Jack says there are more than 49 "nyms" homonyms, holonyms, synonyms, etc.

David Whitten: Jack thinks a focus on a core ontology rather than an upper ontologies.

David Whitten: Jack thinks a domain is an instance of a particular context. Our job is to bridge domains.

Ram D. Sriram: @Gary: I will be going to the Dayton workshop too.

Gary Berg-Cross: @Ram, See your there then.

David Whitten: Jack said that darrett has a four-factor model.

ToddSchneider: Is it not the case that any ontology is developed in one or more contexts? Whether the developer(s) is aware of these contexts and the potential impact of them is not always clear. Attempting to make explicit any 'relevant' contexts is a challenge (e.g., may not be aware of them).

Gary Berg-Cross: @Todd, Competency Qs, used in Ontological Engineering provide an idea of the Scope of an Ontology.

ToddSchneider: Gary, yes. Competency questions provide some qualification of context of an ontology.

ToddSchneider: What relations might a context impact in either the develop or application of ontologies.

Gary Berg-Cross: Context is a recognized Boundary Issue in the Semantic Web. OPDs try to address this he big problem of recognizing and describing the boundaries of a fact, situation, event, from an arbitrary set of data, or from a graph (a description of this problem in the Semantic Web context) See A. Gangemi and V. Presutti. Towards a pattern science for the semantic web. Semantic Web, 1(1-2):6168, 2010.

ToddSchneider: How do contexts relate to a Tarskian view of truth value?

ToddSchneider: Gary, what does 'OPD' refer to?

Gary Berg-Cross: Todd, should be ODP....Ontology Design Pattern.

ToddSchneider: What is a sufficient level of context description for operational purposes?

Gary Berg-Cross: I thought that John Sowa did a good job of outlining some of these issues in http://jfsowa.com/ikl/contexts/contexts.pdf

Jack Ring: Todd, I cited Derrick Cabrera, Cornell, specifically his DSRP model

ToddSchneider: How do competency questions describe context?

ToddSchneider: What 'aspects' of context need to be made explicit in order facilitate interoperability?

janet singer: Questions of context need to be tied to interpretation events for human and/or machine agents. Else there is no hope of moving from an abstract philosophical discussion to engineering capable of improvement IMHO : )

ToddSchneider: Jack, what does 'DSRP' refer to?

ToddSchneider: Janet, could you elaborate?

ToddSchneider: Semiotic events?

Jack Ring: DSRP is Distinctions, Systems (wish he had used Sets), Relations and Perspective www.crlabs.us

Gary Berg-Cross: As good engineering practice designing an ontology requires generic use cases that situate and capture the recurring problems in one or multiple domains. This serves as a first start on addressing an ontology's context by given some idea of scope. We may need to document these better by saying explicitly what is outside of scope.

Gary Berg-Cross: Basic definition of Competency questions - they represent frequent queries that subject matter/domain experts would like to submit to a knowledge base to find answers

Gary Berg-Cross: @janet - Context is also too romantic a term.

janet singer: Interpretations by agents are interpretation of 'texts' meant in a very broad sense any sign event

Gary Berg-Cross: From Sowa slides Contexts in natural languages Literally, a context is text that accompanies a text. More generally, the context may be any background knowledge that helps explain a text.

Gary Berg-Cross: Before I go, we should try to leverage some earlier discussions of context on the Forum. Here is a link to a discussion by Pat Hayes on adding a time context to a 3D object and some of the conceptual and K-rep implications for RDF etc. See http://ontolog.cim3.net/forum/ontolog-forum/2011-02/msg00009.html

Gary Berg-Cross: "here's a non-philosophical way to characterize it. Start with an atomic sentence of the form R(a, b), with no time involved, and suppose that a and b here are ordinary uncontroversial physical objects, say. Intuitively, they are 3D things. Now add time, t. Where do we put the time parameter? Several answers can be given.

  1. Attach it to the sentence, meaning that the sentence R(a,b) is true at the time t. This gives you a hybrid or context logic where the times are possible temporal worlds/indices or contexts, to which truth is relativized. But the sentences being so relativized do not themselves make any reference to time. Call this 3D.
  2. Attach it to the relation as an extra argument, and call the relation a fluent': R(a, b, t) This gives you the classical AI/KR approach which used to be called the situation calculus, where one quantifies over times in the KR language itself, but the object terms are still thought of as denoting 3D rather than 4D entities. Call this 3D+1.
  3. Attach it to the object terms (using a suitable function, written here as an infix @): R(a@t, b@t) This requires us to make sense of this @ operation, and it seems natural to say that it means the t-slice of the thing named, which now has to be re-thought as a 4D entity. So the a, b things have morphed form being 3D (but lasting through time) to being genuinely 4D, and having temporal slices or parts. Call this 4D. For some folk this last step is apparently mind-boggling, although to me it is puzzling how one can think of something being 3D and also extended in time and have it *not* be 4D. For yet other people (think OBO), there are apparently two kinds of thing in the world, one kind (continuants) which must be described using the 3D+1 style , the other (occurrents) which should be described using the 4D style. God alone knows why anyone would believe that there are two ways to exist in time, but there's nowt as queer as folk, as someone's grandmother used to say.

What I like about this way of contrasting the options is that it makes it be simply a matter of syntax - where in the sentence to attach the temporal parameter - and not one of metaphysics. Syntax is way easier than metaphysics. It also means that one can see quite clearly how to make the various formal techniques work together, by allowing the temporal parameter to 'float'. In fact, with a bit of extra work one can embed almost all the necessary temporal reasoning into a generalized unification algorithm which extracts temporal constraints during the unification process. I have all the details somewhere if you (or anyone else) are interested. "

ToddSchneider: Simple question, how do acronyms represent context?

Jack Ring: Looking at Sowa's Diamond I think he noted that 12 aspects of any term or sign must be elucidated in order to minimize misunderstandings

janet singer: A text is that which is interpreted; context is that which impacts interpretation of text by an agent

ToddSchneider: How do (implicit) context(s) impact the (potential) uses of an ontology?

Gary Berg-Cross: Reference on this Grice issue - Bickhard, Mark H. "The social-interactive ontology of language." Ecological Psychology 27.3 (2015): 265-277.

Gary Berg-Cross: I should have thought of Mark Bickhard as a speaker here. I had him and Sowa as speakers on these issue back in a NIST conference 10 years ago or so.

ToddSchneider: How much of the work done in linguistics relative to context is relevant to the development or uses of ontologies (in information systems)?

janet singer: Todd, the challenge is drawing the big lessons from these fields and integrating them in a general (rough) framework

ToddSchneider: How can 'efficacy' of an ontology be captured in requirements for an ontology's expected uses?

ToddSchneider: Is the 'efficacy' of an ontology equivalent to the entailments that can be drawn from it?

janet singer: Ontology can refer to 1) distinctions agents make to act (~ pragmatics), 2) a conceptualization of those distinctions (~ semantics) 3) explicit specification of a conceptualization (~ syntactics)

ToddSchneider: Terry thinks contexts are dynamic.

TerryLongstreth: When I was young I did frequent.........

ToddSchneider: The notion of 'relevance' has embedded context(s).

TerryLongstreth: Janet expressed it well: context has determined and tacit components.

BobbinTeegarden: If you think of ontologies as nesting/holonic, with levels of abstraction, an 'upper' ontology may be decomposed into different contexts in subontologies may seemingly contradict each other ... and this is ok? or do they really conflict at more detail levels?

BobbinTeegarden: Janet: interesting point on explicit vs implicit.

BobbinTeegarden: Jack: the objective is part of the context... yes.

TerryLongstreth: Also, figure-ground relationships have been explored/discussed in art (Escher is an example). We may need to describe context for both.

Cory Casanave: What has clarified for me is the relationship between context, as something that impacts an interpretation, and the interpretation "act". Any such act is "closed" in the set of "text" that impacted the interpretation.

AlexShkotin: C U

Resources

Audio Recording

Synthesis Page

Previous Meetings

... further results

Next Meetings

... further results